Role of Molecular Modeling in the development of Molecular Magnets

Department of Chemistry IIT Bombay

mz=-5 mz=+5ms=-5/2

Small Ueff

Large Ueff

E

ms=+5/2

Isotropic Anisotropic

Prof. Gopalan Rajaraman

1st Nov. IRCC award lecture

Also 8 other papers in this theme: Inorg. Chem. 2015, 1661-1670; Dalton Trans., 2015, 44, 5961-5965; Angew. Chem. Int. Ed., 2014, 53, 2394–2397;Phys. Chem. Chem. Phys., 2014, 16, 14568; Chem. Comm. 2013, 49, 5583.;Polyhedron. 2013, 66, 81-86.; 52, 1299.;Chem. Eur. J. , 2015, 21, 16364.

Theme: Magnetic Exchange in {Gd-3d/radical} Complexes

& Neeraj Tibrewal (Msc Int.)

Magnets vs. Molecule Magnet

[Fe3O4] Magnetic moment arise due to collection of individual magnetic moments of numerous Fe(III) centres. However miniaturisation demands permanent magnetization in a as tiny space as possible.

Applications of magnets widespread, information Storage devices using magnetic materials is a Multi-billion dollar industry.

Is it possible to obtain detectable and measurable permanent magnetization in a molecule?

What are Molecule Nano Magnets?

A single molecule that behaves as a nanoscale magnet below a critical temperature. i.e. displays hysteresis of molecular origin.

Interests in single molecule magnets As single molecule magnetic memory devices As magnetic quantum logic devices Fundamental studies of large spins, i.e. quantum vs. classical behavior Advantages Chemical control - "bottom-up" materials engineering Tremendous control of the magnetic unit (the spin), as well as its coupling to the environment

Why do we need Molecular magnets? Evolution of areal density in magnetic data storage

Kryder’s law: magnetic disk areal storage density doubles annually

SMM on surfaces

RAMAC- 20 MB

Superparmagnetic effect

35 billion xincrease

GMR

Potential applications

7

e-

µorbital

µspin

µtotal

How do they work? Employing Lanthanides for molecule based magnets is an attractive idea.

Why Lanthanides are so attractive ? Presence of unquenched angular momentum

Strong spin-orbit coupling (LS coupling)

State-of-the art and challenges:

[(CpiPr3)2Dy]+ exhibit barrier height

grater than 1800 K but blocking

temperature is 60 K while

[(CpiPr5)Dy(Cp*)] exhibit 80 K

Very fast quantum tunnelling !

Chilton & Miles, et al.,

Nature., 2017,548, 439.

Layfield et al.,

Science, DOI:10.1126/science.aav0652.

mJ=-15/2 mJ=+15/2

mJ=-13/2 mJ=+13/2

mJ=0

[Dy]

1800K

E

Understanding and controlling spin Hamiltonian parameters is important for future success.

Computational Chemistry can play an important role here !

Our approach:

Methodology: Model: Broken symmetry Model described by Noodleman Software: Gaussian 03/09 and Jaguar Using: B3LYP Functional and TZV basis set Relativistic corrections by DKH or ECP Also utilized ab initio CASSCF calculations (MOLCAS, ORCA).

How do we stop tunnelling? For tunneling to occur between two levels that have the same energy, some admixing of the states must occur

Interacting states giving rise to a tunnel splitting

This interaction can be provided by: low-symmetry components of the crystal field by hyperfine fields (provided by magnetic nuclei e.g. 55Mn) by dipolar fields caused by metal ions.

mJ = -15/2

mJ = +15/2

MJ

Can we deliberatively include 3d metal ion with Lanthanides to Stop tunnelling?

Obtaining large exchange interaction J suppress the QTM

Weak exchange Stronger exchange

E

Ln Ni

Other issues:

QTM

{3d-4f} molecular Assembly quench tunnelling

E(S=0) – E(S=1) _______________ 2S1S2 + S2 ST= 5/2

ST= 9/2 -J

ST= 5/2

ST=9/2

+J

What is S and J ?

Estimating Super-exchange using DFT

Gd Ni

S= 7/2 S= 1

J =

Gd Ni

S= 7/2 S= -1

+J – Ferromagnetic coupling -J – Anti-ferromagnetic coupling

A periodic walk on mixed {3d-Gd} complexes (M=VIV,FeII, CoII, NiII and CuII )

Complexes Jcal Jexp

[L1VO(H2O)]

+1.5

+2.2

[L1Fe(CH3OH)Gd(NO3)3] +1.1 +0.5

[L2Co(MeOH)Gd(NO3)3] +0.9 +0.9

[LNi(H2O)2Gd(NO3)3] +2.1 +3.6

[(NiL)Gd(hfac)2(EtOH)] +0.3 +0.3

L1=1,3-diamino-2,29-dimethylpropane L2=N,N9bis

(3-methoxysalicylidene)-1,2-diamino-2-methyl

propane L2=1,3-bis[(3- methoxysalicylidene) amino]-

2,2-dimethylpropane.

J value estimation from DFT in {3d-4f} clusters

Costes et al., Inorg. Chem. 2000, 169.

Gd 3d

O N

C

Linear M-O-M angle leads to Overlap od 3d orbitals and hence antiferromagnetic coupling

Unusual Magnetic coupling in {Gd-3d/radical pairs}

Limiting the M-O-M angle to 90 deg. leads to zero overlap and ferromagnetic coupling

General {3d-3d} interactions:

Gd Ni

O

Gd Ni O Always Ferromagnetic !

Ab-initio calculations on VIV-GdIII pair: Role of vacant 5d orbital of Gd

3d (V)

4f(Gd)

5d(Gd)

Bridging ligand

6.99 α 0.98 α

0.02α

0.01 β

Methodology Exchange

Coupling

J (cm-1)

DFT +2.3

CASSCF (8,8)

(f7+d1)

+1.1

CASSCF (8,9)

(f7+d1+6s0)

+1.1

CASSCF (8,13)

(f7+d1+5d0)

+1.3

CASSCF (8,14)

(f7+d1+5d0+6s0)

+1.4

Experimental +1.5

ORCA code Gd SARC, rest TZV basis set Relativistic: ZORA approach

General Mechanism magnetic coupling

G d (II I ) 4 f

G d (I I I ) 5 d

G d (II I ) 4 f p o lr i .

TM

TM

~ 1 0 0 k J /m o l

Very little ligand mixing – Moderate Js achievable

3d-4f

80 85 90 95 100 105 110 115 120 125 130

-2

0

2

4

6

{Fe-Gd}

{Co-Gd}

{Ni-Gd}

{V-Gd}

J (c

m-1)

Bond Angle

Magneto-structural correlations

0 5 10 15 20 25 30

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

{Co-Gd}

{Fe-Gd}

{V-Gd}

J (c

m-1

)

M-O-Gd-O Dihedral Angle

{NiGd}

2.0 2.1 2.2 2.3 2.4 2.5 2.6

0.8

1.6

2.4

3.2

{Fe-Gd}

{Co-Gd}

{V-Gd}

{Ni-Gd}

J (c

m-1)

avg M-O Bond distance

Gd

3 d G d 5 d

M

Gd

3 d G d 5 d

MJ (

cm-1)

V-O-Gd-O dihedral angle

V-O

-Gd

bon

d a

ngle

Large J to quench QTM not possible for {3d-4f} pairs

So - dead-end ?

Stronger overlap enable stronger coupling

[{[(Me3Si)2N]2Gd(THF)}2(μ-η2:η2-N2)]-

An extremely strong AF exchange between radical and GdIII has been observed.

N23- Exp : JGd-rad = -27 cm-1

DFT: JGd-rad = -23 cm-1 JGdGd= -0.5cm-1

J. R. Long et al. Nat. Chem., 2011, 37, 538.

BUT Why ???

The unpaired electron on N23- resides in a * orbital.

Large negative charge and diffused nature of * orbital leads to strong overlap particularly with the fxyz orbital of GdIII

Jnet=JF+JAF Small & weak Large & strong

Exchange as high as 40 cm-1 is achievable.

How do we make it even stronger?

Several fullerene and radical fullerene are screened for a strong exchange

interactions.

Gd2@C79N radical cage:

DFT: JGd-rad = +378 cm-1

JGdGd = 0.8 cm-1

Significant CT leads to largest J ever observed for this cage.

Exp evidence: Significant CT and long relaxation time and EPR experiments indicate strong coupling.

H. C. Dorn, et al. J. Am. Chem. Soc., 2011, 133, 9741.

Exchange coupling: B3LYP/TZVP

J = 340 ±10 cm-1

Radical Rays of hope !

Summary: Molecular modelling is powerful tool in the area of MNMs {4f-4f} pairs mediate weak exchange ( maximum 2 cm-1) Quenching of QTM is challenging. Molecular coolants is the ideal application. {3d-4f} pairs mediate moderately strong exchange (up to 10 cm-1}. QTM suppression is possible, new structures and avenues needs to be explored. {radical-4f} pairs mediate strong exchange (up to 400 cm-1). Less explored area and ideal for SMMs

Computing facility HPC Facility - IIT Bombay

IIT-Bombay

SERB